Some Longines Repair Questions

[gmorse]

Hi John,

Let me know what I've gotten wrong!

It all looks fine, just a smear of the grease in the barrel, and be very careful where you apply the 9415, the BHI document does describe all this. There is some debate about whether the lever/pallet arbor pivots should be lubricated at all; they certainly shouldn't be in wristwatch sized movements but opinions vary on pocket watches, I'd be inclined to lubricate them sparingly since they're only moving through a small arc.

Regards,

Graham

 

[John Hinkey]

The BHI lubrication guide is very useful. https://theindex.nawcc.org/Articles/BTI-The_Practical_Lubrication_of_Clocks_and_Watches.pdf

The commonest beginner mistake is too much oil and I have made that mistake. You need good light and magnification to place the oil successfully. If you mess it up, take it apart, swish in degreaser and try again. The pallet stone (exit pallet) can be oiled through the inspection holes in the plate with light from underneath (a torch will do).

Make sure you put the new spring in the right way up. I have to check repeatedly because for some reason it confuses me. Check the spring coils are coiled in the correct direction so that the barrel arbor hook will engage. Before you fit the spring it would be a good idea to check the assembled barrel for side shake and end shake as this can seriously affect power delivery.

Before any oiling check the train runs freely. Of course be extremely careful that the pivots are in place before tightening bridges. Small springs in the keyless works can try to escape so restrain them with peg wood or a plastic probe as you are fitting them (and/or work in a plastic bag if in doubt).

Thanks for the advice - the mainspring install is the biggest challenge it seems, but I took copious pictures of the spring in the barrel while taking it apart and I have the still-in-the-barrel spring from the donor watch to look at. I re-assembled the train on the donor watch and got a feel for how to get all the pivots in their right place before installing any screws.

I'll let everyone know how it goes.

- J

 

[John Hinkey]

Hi John,


It all looks fine, just a smear of the grease in the barrel, and be very careful where you apply the 9415, the BHI document does describe all this. There is some debate about whether the lever/pallet arbor pivots should be lubricated at all; they certainly shouldn't be in wristwatch sized movements but opinions vary on pocket watches, I'd be inclined to lubricate them sparingly since they're only moving through a small arc.

Regards,

Graham

Thanks Graham!

 

[John Hinkey]

Ah, my first issue is that the mainspring T end ears will not fit into the holes in my barrel:

so I will have to file/grind down the ears that stick out to fit the holes in the barrel, just like I saw in the (original?) spring I took out:

Ah more fun . . .

- J

 

[John Hinkey]

Mainspring T-end modified and hand wound in. That was not the most fun thing to do, but I think I did a decent job of it.
Pretty sure I got the direction correct:
Original:

New:

Fully assembled and ready for installation:

I did put three little radial smears of KT-22 on the bottom before putting in the spring and on top of the spring before putting the lid on.

I then used the winding stem, which fits the square on the arbor, to hand wind the spring about a turn or so and it felt quite free and torquey and smooth.

Next up is re-assembly of the balance (putting the spring back on and oiling the balance cock jewel) and then initial trial assembly to see if it wants to run with a bit of power put into the gear train.

- J

 

[John Hinkey]

And those marks distributed on the side of the spring - those are not debris as they were there right out of the packet. They appear to be slight cosmetic nicks in the side. It feels smooth to the touch. Perhaps I should buy another from a different vendor in case this one causes trouble . . .

 

[svenedin]

And those marks distributed on the side of the spring - those are not debris as they were there right out of the packet. They appear to be slight cosmetic nicks in the side. It feels smooth to the touch. Perhaps I should buy another from a different vendor in case this one causes trouble . . .

I’d suggest leaving it for now. Well done on getting the spring installed. I’d also suggest leaving the train un-oiled for now whilst you reassemble and test. If you encounter problems it makes a mess taking apart a freshly oiled watch (although if you want the practice go ahead). I wear disposable gloves to re-assemble (powder free).

 

[John Hinkey]

I’d suggest leaving it for now. Well done on getting the spring installed. I’d also suggest leaving the train un-oiled for now whilst you reassemble and test. If you encounter problems it makes a mess taking apart a freshly oiled watch (although if you want the practice go ahead). I wear disposable gloves to re-assemble (powder free).

I think I will take your advice and re-assemble un-oiled. I will oil the balance capstones as those are somewhat of a pain (especially on the balance cock). This movement is fairly easy to take apart and re-assemble so it shouldn't be too much of a hassle to disassemble and oil. I've already practiced removing the mainspring barrel and bridge and re-installing it on one of my donor movements and know how tricky it is to get all the wheels in place and pivots in the right holes of the bridge before installing any screws.

Thanks!

- John

 

[karlmansson]

I would like to bring to attention that this thread is now at 158 posts, this not included. While many ask similar questions to yours, John, few receive the same degree of response and walk through style guidance.

I gather two things from this that I would encourage others to copy: provide good images! Yours have been stellar John!
Provide detailed descriptions of the problem and adopt nomenclature when you learn it. Providing feedback on how things go/went will also leave those who help feeling engaged and appreciated for their help (at least I feel that way). Sometimes there are “drive by” posters that drop a question and never show up again. Doesn’t inspire a lot of goodwill.

Good on ya, John! Hope you get your watch to run!

Regards
Karl

 

[John Hinkey]

Thanks for the compliments. Being an engineer and a photographer certainly helps with the process and the photo gear.

The reassembly will have to wait until next week as I'm currently flying to the East Coast for the weekend.

 

[John Hinkey]

Ah, getting it back together.

When I spin the barrel everything in the train runs smoothly it seems.
However, I found that my click spring is bent (teardown pic) and I need to bend it back gently:

then I can install the pallet fork and put some power into the mainspring to see if the fork snaps back and forth.
If so then I will re-assemble and install the balance to see if it runs at all (un-oiled).

Small steps.

- John

 

[John Hinkey]

Well, it's basically together - I fixed the click spring and installed it so I can wind up the mainspring using the square key:

Had to put the balance back together which means oiling the cap jewels:

And had to re-install the balance onto the cock and get it installed.
And . . . . it started running right away as I had just put a half-turn into the spring to power it and it ran for an hour or so by itself:

So I need to oil it and install the winding mechanisms so I can do some timegraphing to see how bad it is.

I was shocked when it just started running as soon as the balance was in!

 

[gmorse]

Hi John,

I was shocked when it just started running as soon as the balance was in!

It's always a relief when this happens, but it shows that your efforts have been in the right direction. I think you should be very pleased with yourself!

Regards,

Graham

 

[John Hinkey]

Thanks Graham. These natural ruby cap jewels make it hard to see where the oil went after the cap is installed due to all the inclusions.
I had to remove the balance and adjust (rotate) the pin that holds the hairspring to the balance cock in order to have the coils of the balance spring be concentric and uniformly spaced:

ah, that's better:

here's what my initial installation looked like and the coil was not quite right:

On to pivot oiling and installing the winding works so I can fully wind it and put it on the timegrapher to see how it goes.
This movement probably hasn't worked in 100 years since its owner died in 1890 and it's been knocking around in boxes and drawers since then.

More later today.

- John

 

[John Hinkey]

Here's a short video of the movement running initially after the first assembly with just the first couple wheels lubricated as well as the balance pivots.
Only about a 1/2 turn on the barrel to power it.

 

[Skutt50]

I had to remove the balance and adjust (rotate) the pin that holds the hairspring to the balance cock in order to have the coils of the balance spring be concentric and uniformly spaced

That could be what was needed however you might want to verify that the coil stays concentric (balance at rest with the mainspring let down) if you move the regulator to its end positions. During this test the coil should stay in the middle of the "pins" of the regulator and not touch any "pin". If the hairspring is not in the middle all the way you may find some odd behavior when regulating the movement, e.g. if you try to speed it up by moving the regulator the movement slows down........

 

[John Hinkey]

Anyone know what the lift angle of this movement should be? I will try to measure it, but if I don't have to that would be quite appreciated.
So far I get a beat error of around 2msec and an amplitude around 200 deg. with a 52 deg. assumed lift angle (which is probably too high).

I wound it up about half way and let it run overnight and it was still going strong this morning.

Thanks -

John

 

[gmorse]

Hi John,

Anyone know what the lift angle of this movement should be? I will try to measure it, but if I don't have to that would be quite appreciated.

It's probably in the 40s rather than the 50s. The value is only used by the timer to calculate the amplitude and there are other ways to determine that.

Regards,

Graham

 

[John Hinkey]

Hi John,


It's probably in the 40s rather than the 50s. The value is only used by the timer to calculate the amplitude and there are other ways to determine that.

Regards,

Graham

Well, using the method of marking the 180 deg. point on the balance wheel and winding it just enough to get 180 deg. amplitude I get about 37 degrees on the timegrapher to get 180 deg. amplitude.
I did see a post that says Longines pocket watches should be about 35 deg lift angle:

https://mb.nawcc.org/threads/pocket-watch-lift-angle.173360/

- J

 

[John Hinkey]

Well, when I set the timegrapher to 37 degrees lift angle and fully wind the watch I get:

• About 150 deg. amplitude
• 2.0 to 1.8msec beat error
• +35 to -35 s/d variation in rate over 5 minutes or so time period.

EDIT: These numbers are at dial down position

I will admit that I was very conservative on oiling the wheel jewels - pretty much all the oil I put in the cup sucked right in to the pivot-jewel radial gap and very little stayed in the cup - did I not oil this enough? I was also not over the top on the cap jewels for oil based on what I've seen on the internet for the amount to use.

- J

 

[gmorse]

Hi John,

I will admit that I was very conservative on oiling the wheel jewels - pretty much all the oil I put in the cup sucked right in to the pivot-jewel radial gap and very little stayed in the cup - did I not oil this enough? I was also not over the top on the cap jewels for oil based on what I've seen on the internet for the amount to use.

I don't think the amount of oil you used, (and you did indeed use some), would account for the figures you quote. The low amplitude and fluctuating rate are symptoms of some structural problems, possibly wear in pivots or their holes. The beat error could be adjusted closer but it isn't a big issue.

Is there a consistent frequency to the rate fluctuations? Do you get comparable numbers with it dial up and in vertical positions? Are you measuring with the dial and hands, (and hence the motion work), on the movement?

Regards,

Graham

 

[John Hinkey]

Hi John,


I don't think the amount of oil you used, (and you did indeed use some), would account for the figures you quote. The low amplitude and fluctuating rate are symptoms of some structural problems, possibly wear in pivots or their holes. The beat error could be adjusted closer but it isn't a big issue.

Is there a consistent frequency to the rate fluctuations? Do you get comparable numbers with it dial up and in vertical positions? Are you measuring with the dial and hands, (and hence the motion work), on the movement?

Regards,

Graham

Dial down position only. No dial or hands installed. Motion works installed.
The rate fluctuations (sinusoidal in nature) are on the order of handful of minutes in period. Shall I try it on other orientations to see what happens or should I install the dial and hands and see what time deviation I get over a 24 hour period?

I also inspected the roller jewel when the balance is stationary (watch powered down) and it's very well centered between the banking pins. One of the banking pins does seem to be very slightly bent outwards though.

Thanks -

John

 

[John Hinkey]

Remember that this watch has:

• A non-original replacement mainspring that was about 1-2cm longer than the original
• A cracked jewel on the Third Wheel
• A forth wheel with a very slightly bent pivot that the seconds hand attaches to (noted above, but I did not try to fix it)

I have not demagnetized the movement.

Any thoughts of what to do? It's easy enough to fully take apart and re-inspect things.

 

[John Hinkey]

Also, to check the actual amplitude I took a 120 frame/second video of the balance with it not-quite fully wound (it had wound down a few hours from fully wound) and a mark on the rim to locate the 180 deg. amplitude point:

and here's my estimate of where the black mark on the rim of the balance is on the opposite oscillation/swing of the balance wheel:

so it looks like the I'm getting ~250 deg. of amplitude in actual and it's within about 10 deg. of being symmetric (which is likely within the accuracy of my video which really needed to be 240 frames per second frame rate).

So perhaps my lift angle estimate is off by quite a bit as I was looking for this rim mark by eye @ the 180 degree mark which was quite hard to do.
I also noticed that over the length of a 15 second or so video the rim mark comes to stop at the same spot/location to within the width of the mark on the rim for all the oscillations - not bad I suppose?

Other things to notice is how close the coils are getting to each other in the first image:

wish I could post an embedded video that is way slowed down . . .

- John

 

[John Hinkey]

Update, I just took 120 fps video to confirm ~180 deg. of rotational amplitude at low power and I need to set the lift angle to 58 deg. on the timegrapher to have it display this amplitude value. This seems weird.

BUT - I also took 120fps video nearly fully wound with the timegrapher telling me about ~240 deg. of amplitude and what I see in the video is very consistent with my video frame grab images above (~250 deg. of amplitude).

SO - it appears that this particular Longines movement has a lift angle of ~58 deg (or thereabouts) @ 180 deg. of amplitude and an actual full power amplitude of 240 deg. or so. Gradual rate variation of +/-20 sec/day over a few minute period. Still getting about 1.9msec beat error.

How would my findings be affected by something out of sorts in the mechanism?

• Should I demagnetize the movement?
• Adjust the balance spring to prevent it from getting so close to each other?
• Or should I make a disassembly to see how the oiling of the first couple of wheels is going and any contamination might have been done during my assembly?
• Straighten the fourth wheel dial side pivot?
• Straighten the banking pins slightly?
• Should I call this good enough for a cracked-jewel rebuilt by an amateur 140 year old pocket watch movement?

- John

 

[gmorse]

Hi John,

1. Should I demagnetize the movement?
2. Adjust the balance spring to prevent it from getting so close to each other?
3. Or should I make a disassembly to see how the oiling of the first couple of wheels is going and any contamination might have been done during my assembly?
4. Straighten the fourth wheel dial side pivot?
5. Straighten the banking pins slightly?
6. Should I call this good enough for a cracked-jewel rebuilt by an amateur 140 year old pocket watch movement?

1. Yes.
2. Yes, it doesn't appear to be properly centred; take the spring off the balance and fit it to the cock by itself to check if the collet is resting exactly over the jewel hole. An off-centre spring will be putting unwanted sideways pressure on the balance.
3. The amount of oil shouldn't make much difference, it should run without any at all just as a test, but contamination is worth checking for.
4. Ideally, the bent pivot should be addressed, but there's a risk that it could break if you don't have the equipment to do it safely.
5. The banking pins should be upright, they were planted at the factory in the correct position, but the attentions of previous repairers will have confused the picture.
6. That's up to you, but using this as a learning tool has been productive, so I would continue to investigate what's causing the issues even if you can't at the moment remedy all of them.

Regards,

Graham

 

[John Hinkey]

Hi John,

1. Yes.
2. Yes, it doesn't appear to be properly centred; take the spring off the balance and fit it to the cock by itself to check if the collet is resting exactly over the jewel hole. An off-centre spring will be putting unwanted sideways pressure on the balance.
3. The amount of oil shouldn't make much difference, it should run without any at all just as a test, but contamination is worth checking for.
4. Ideally, the bent pivot should be addressed, but there's a risk that it could break if you don't have the equipment to do it safely.
5. The banking pins should be upright, they were planted at the factory in the correct position, but the attentions of previous repairers will have confused the picture.
6. That's up to you, but using this as a learning tool has been productive, so I would continue to investigate what's causing the issues even if you can't at the moment remedy all of them.

Regards,

Graham

As usual, thanks Graham!

I also forgot that this balance has a slightly tilted impulse jewel to go along with all the other issues.
I will demagnatize it and partially disassemble it to make sure there is no debris in the jewels. I'll also inspect the balance spring as you described.
In the mean time I will be working on getting the silver hunter case restored.

Can one post videos in the forum that aren't attachments, but rather play directly in the post?

- John

 

[gmorse]

Hi John,

Can one post videos in the forum that aren't attachments, but rather play directly in the post?

Yes, but it seems that linking to a video hosting site such as Vimeo or YouTube is more faster and more efficient.

Regards,

Graham

 

[John Hinkey]

Hi John,

1. Yes.
2. Yes, it doesn't appear to be properly centred; take the spring off the balance and fit it to the cock by itself to check if the collet is resting exactly over the jewel hole. An off-centre spring will be putting unwanted sideways pressure on the balance.
3. The amount of oil shouldn't make much difference, it should run without any at all just as a test, but contamination is worth checking for.
4. Ideally, the bent pivot should be addressed, but there's a risk that it could break if you don't have the equipment to do it safely.
5. The banking pins should be upright, they were planted at the factory in the correct position, but the attentions of previous repairers will have confused the picture.
6. That's up to you, but using this as a learning tool has been productive, so I would continue to investigate what's causing the issues even if you can't at the moment remedy all of them.

Regards,

Graham

This kind of curve is exactly what I'm seeing on my Timegrapher:

wavy wandering over a long time period AND when I flip the movement over things drastically change.

Cracked and Chipped Jewels as Causes of Unstable Rates and Wavy Timing Machine Traces • Adjusting Vintage Watches

This post describes why cracked and chiped jewels can cause a watch to have unstable rates and wavy timing machine traces.

adjustingvintagewatches.com

which the author says is typically due to cracked/chipped jewels, which I definitely have one cracked jewel on the third wheel.

I will do all the other stuff you noted, but I suspect the cracked jewel is my primary problem, which, unfortunately, is the hardest one to fix (I do have a scrap movement with a good jewel for this cracked one, but this is outside my current skill set) and won't happen any time soon.

More later tonight.

- J

 

[John Hinkey]

Well then, I tried the easy thing of de-magnetizing the watch and wow did it make a difference!
The watch went from +/-40 s/day swings to +/-5 or so
I used one of these cheap ones doing 4 demangetizing cycles:

How To Demagnetize Your Watch

What to do when your watch is too attractive.

www.hodinkee.com

so it appears I had something magnetized in the movement. I will do a few more demagnatizing cycles just to be sure.

Anyways, I can get the rate to be +10 to +15 s/day and roam around in this range over a long time period at full power/full wind. The regulator on the balance cock is extremely sensitive to movements/adjustments so it's hard to get it near zero s/day rate.

Progress. If it stays in this kind of accuracy range I may call that good enough and do a full assembly with the dial and move on to my next heirloom pocket watch (Tavannes 17 jewel open face ~1920's vintage).

Thanks Graham!

EDIT: Graham I will remove the balance and see what's going on with it and inspect the pivot jewels to see how much oil is in the cups.
I don't think I'll do anything with the banking pins or impulse jewel at this point (and replace the cracked jewel later too).

 

[svenedin]

It is very easy for parts to get magnetised. I use non-magnetic tweezers but my screwdrivers are steel. I demagnetise before starting work but often notice my screwdrivers become magnetised after just a few hours. The last watch I worked on developed a magnetised hairspring. I thought I had damaged it (coils not concentric) but it was just magnetised.

 

[John Hinkey]

It is very easy for parts to get magnetised. I use non-magnetic tweezers but my screwdrivers are steel. I demagnetise before starting work but often notice my screwdrivers become magnetised after just a few hours. The last watch I worked on developed a magnetised hairspring. I thought I had damaged it (coils not concentric) but it was just magnetised.

Yeah, I found my screwdrivers to be very magnetized and the cheap magnetizer/demagnetizer plastic coated magnets couldn't do the job. So I bought the electric one above and it did the job nicely.

I just took a look at some of the pivots in the jewels and I cannot seen oil in the gaps on some of them so I will add just a little more since I was very conservative the first time through. It also appears that my escape wheel bottom pivot is every so slightly bent too.

I keep having to remind myself it's a 140 year old watch and won't be perfect . . .

 

[John Hinkey]

Well, I felt good so I went and very very very very carefully bent the offending banking pins back to vertical and . . . it would not run.
I looked at the pallet fork and escape wheel and it appeared the the un-bent banking pin (which was still slightly bent outward) did not give the pallet enough angular freedom for the escapement to work.
Pulled the balance and slightly bent both pins back, but not quite to where it was originally and now it runs.

BUT now the beat error is ~3msec whereas before my fix it was 1.8 or so.
Related to this, I found that the impulse jewel is slightly rotated when the balance as at rest towards the bent banking pin.

By my estimation someone bent the one banking pin outward slightly to compensate for the impulse jewel being mis-aligned.
When I bent the banking pin inward to be truly vertical it aggravated the mis-aligned impulse jewel location.

So you can in fact make things worse by "fixing" things.

As Graham noted, what I need to do now is rotate the balance spring so that the impulse jewel is lined up properly to get the beat error lower.

In this state in the dial down position it's stable at +10s/day +/-5 wandering and goes to +45 sec/day in dial up position.

 

[John Hinkey]

Now, I went and adjusted the mounting stud angle such that the impulse jewel was aligned properly and the beat error came down to 1.0msec - even better than before my banking pin "fix". Now it's running at 0s/day +/-5 sec/day or so in Dial Down position.

What did I learn? ANYTHING that upsets the pallet fork/impulse jewel alignment is bad and that just because something like a banking pin on an old watch looks a bit wonky, it doesn't necessarily mean it should be fixed (or, be prepared to fix something else after you've solved one problem because sometimes two problems can sort of cancel each other out).

So now at 58 deg. lift angle, I'm getting 240 deg. of amplitude, 1msec beat error, and stable operation around 0 s/day (+/-5 s/day) fully wound in the dial down position.
If I simply rotate it to the dial up position, beat error goes to 1.2msec, rate of +20sec/day(+/-5) and the amplitude is 235 deg. or so.

I'm going to leave it as is and solve it's other problems another day.

You all have been invaluable in my journey to learn this stuff!

Hopefully my Great Uncle is smiling somewhere because his great nephew is getting his watch running again.

- John

 

[karlmansson]

Bent bankings are usually a sign that someone before you has tried to compensate for another problem. As you have probably noticed of figured out, moving the bankings will alter the lift angle. Moving the pallet stones in or out will alter the lock and draw on the escape wheel but not the lift angle. Some repairers attempt to alter the locks by moving the bankings instead, which can be done, not realizing that they will also have messed up the lift angle. Bankings should, as a rule, be straight and parallell. They were made that way from the factory. Some American pocketwatches will have bankings that can be rotated though, so some means of adjustment is provided sometimes.

Glad you got it running!

Regards
K

 

[John Hinkey]

Bent bankings are usually a sign that someone before you has tried to compensate for another problem. As you have probably noticed of figured out, moving the bankings will alter the lift angle. Moving the pallet stones in or out will alter the lock and draw on the escape wheel but not the lift angle. Some repairers attempt to alter the locks by moving the bankings instead, which can be done, not realizing that they will also have messed up the lift angle. Bankings should, as a rule, be straight and parallell. They were made that way from the factory. Some American pocketwatches will have bankings that can be rotated though, so some means of adjustment is provided sometimes.

Glad you got it running!

Regards
K

Thanks - It just finished running for about 36 hours on a full wind. I have yet to put the dial back on with the hands to see how well it keeps time fully assembled.

 

[roughbarked]

Repeat alternatively: If it ain't broke, don't fix it. If it ain't fixed, don't break it.

 

[John Hinkey]

Repeat alternatively: If it ain't broke, don't fix it. If it ain't fixed, don't break it.

Sometimes it's hard to figure out if something is really broken or if it was really fixed

 

[roughbarked]

Sometimes it's hard to figure out if something is really broken or if it was really fixed

Life wasn't meant to be easy. However, you now have more knowledge of this than you had before.